Apple has released the M5 Pro and M5 Max chips alongside a new MacBook Pro update, and both processors represent a more significant architectural shift than past generational upgrades typically deliver.
Normally, Apple’s Pro and Max variants take the base chip’s core design and scale it up: more CPU cores, more GPU cores, more memory bandwidth. The M5 generation does something different, both in how the silicon is physically constructed and in the types of CPU cores it uses.
A New Multi-Chiplet Design
Apple describes the M5 Pro and M5 Max as using an “all-new Fusion Architecture” that combines two silicon chiplets into a single processor. Apple has used multi-die packaging before, but only to merge two Max chips into an Ultra. This is different. Rather than doubling a base chip, Apple has split responsibilities across two dies built on TSMC’s 3nm process.
The first die handles the CPU, Neural Engine, SSD controller, Thunderbolt, and display output. This die is identical across both the M5 Pro and M5 Max. The second die is where the two chips diverge.
- The M5 Pro gets up to 20 GPU cores, one media engine, and up to 307 GB/s of memory bandwidth.
- The M5 Max gets up to 40 GPU cores, two media engines, and up to 614 GB/s of memory bandwidth.
The doubling of every figure in the M5 Max’s GPU die suggests Apple is combining two M5 Pro GPU dies to produce the Max configuration, a clean architectural split that keeps the CPU side unified.
Three Types of CPU Cores
Apple’s spec sheets now list three distinct CPU core types: “super” cores, performance cores, and efficiency cores. This is a meaningful change from prior generations, which used only two.
“Super cores” is Apple’s new label for what it previously called performance cores. That rebrand applies retroactively to the standard M5 as well. The efficiency cores remain at the bottom of the hierarchy, six in total, unchanged in name and still tuned for low power draw.
The genuinely new addition sits in the middle: a third core type called “performance cores,” used exclusively in the M5 Pro and M5 Max. These cores share design characteristics with the super cores but prioritize multi-threaded throughput over peak single-core speed. The approach draws a conceptual parallel to AMD’s use of standard and compact Zen cores in its laptop processors, where larger cores chase peak clock speeds and smaller variants optimize for die space and thermal efficiency.
The M5 Pro and M5 Max each carry 18 CPU cores total: 6 super cores and 12 performance cores. By comparison, the M4 Pro offered up to 10 large cores and the M4 Max up to 12. On paper, the raw count of high-performance cores has dropped, but Apple expects the higher single-thread capability of the super cores combined with the 12 performance cores to produce a net gain in real-world throughput.
Performance Unknowns Remain
Whether the architectural changes translate into meaningful performance gains over the M4 generation won’t be clear until independent testing begins. The structural changes are real and well-documented. Their practical impact is not yet measured.
Photo by BoliviaInteligente on Unsplash
This article is a curated summary based on third-party sources. Source: Read the original article
